Charge for an SDSS Director’s Review of the
March 1-2, 2004
This charge is for the Committee to conduct an SDSS Director’s Review of the Catalog Archive Server (CAS) effort at Johns Hopkins University (JHU) and Fermilab. The CAS provides the SDSS collaboration and the general astronomy community with an interface to the SDSS data products. The CAS is being developed at JHU; the current plan is to load and deploy the production CAS from Fermilab. This review will focus on the remaining development work for the CAS, the computer hardware required to deploy the CAS, and the plans for implementing the CAS at Fermilab.
The goal of the review is to assure SDSS management, as well as JHU and Fermilab management, that the plans and resources for completing the CAS and implementing it at Fermilab are reasonable and in place, or at a minimum that a reasonable plan and schedule exists for obtaining resources not yet in place (e.g., computer hardware systems).
Therefore, review presentations will include discussions of the SDSS data distribution plan; CAS development efforts, current status and remaining work; computer hardware requirements and plans for deploying and supporting the production CAS; plans for integrating the CAS into the production environment at Fermilab; and the current status of integration and lessons learned to date. The Committee is requested to comment on the information and plans presented during the review, with particular focus on the following questions:
Are the requirements for the CAS well-defined? If not, what areas should be clarified?
Is there a clear path to completion of the CAS development?
Is the hardware implementation plan reasonable? In particular, is the design of hot spares, the types of RAID arrays, etc. appropriate for meeting the requirements? Is the plan for scaling the CAS to meet the needs at the end of the survey sensible? Are the hardware costs well understood?
Are the resources, including manpower at JHU and Fermilab, to deploy and operate the CAS reasonable? Is the level of interaction between JHU and Fermilab sufficient?
The Committee should also assess how well the SDSS team has addressed recommendations made in the July 20-21, 2000 Director’s Review of SDSS Data Processing and Distribution that are relevant to the scope of this review.
Finally, the Committee should present findings, recommendations, comments and conclusions at a closeout meeting with SDSS management and the review participants, and provide a written report to the SDSS Director soon after the review.
Review of CAS Development and Deployment Effort
March 1-2, 2004